Re: [CWM] Trignometric Module

I just updated and tried to run the regression tests
and I got:

$ make
rm ../*.pyc
touch pyc-check
PYTHONPATH=`/bin/pwd`/.. python retest.py -c regression.n3
list/detailed.tests
online.tests norm/detailed.tests
Traceback (most recent call last):
  File "retest.py", line 274, in ?
    main()
  File "retest.py", line 162, in main
    kb = loadMany(testFiles)
  File "/home/connolly/w3ccvs/WWW/2000/10/swap/myStore.py", line 98, in
loadMany
    return _checkStore().loadMany(uris)
  File "/home/connolly/w3ccvs/WWW/2000/10/swap/myStore.py", line 45, in
_checkStore
    store = storeClass() # Make new one
  File "/home/connolly/w3ccvs/WWW/2000/10/swap/llyn.py", line 1150, in
__init__    import cwm_trigo   # Trignometry
  File "/home/connolly/w3ccvs/WWW/2000/10/swap/cwm_trigo.py", line 96
    def evaluateSubject(self, obj_py):
                                      ^
IndentationError: unindent does not match any outer indentation level

that was easy enough to fix... (checked in 1.6) but then I tried
pychecker (http://pychecker.sourceforge.net/) on it:

$ pychecker cwm_trigo.py
Processing cwm_trigo...
 
Warnings...

[... stuff about term.py elided ...]

cwm_trigo.py:19: Variable (__cvsid__) not used
cwm_trigo.py:22: Imported module (math) not used
cwm_trigo.py:22: Imported module (re) not used
cwm_trigo.py:22: Imported module (urllib) not used
cwm_trigo.py:32: Module (sys) re-imported
cwm_trigo.py:70: No global (acos) found
cwm_trigo.py:74: No global (asin) found
cwm_trigo.py:78: No global (atan) found
cwm_trigo.py:83: No global (atan2) found
cwm_trigo.py:84: Function returns a value and also implicitly returns
None
cwm_trigo.py:87: No global (cos) found
cwm_trigo.py:91: No global (cosh) found
cwm_trigo.py:95: No global (degrees) found
cwm_trigo.py:97: No global (radians) found
cwm_trigo.py:101: No global (sin) found
cwm_trigo.py:105: No global (sinh) found
cwm_trigo.py:109: No global (tan) found
cwm_trigo.py:113: No global (tanh) found

This code doesn't look like this code can work. I think you
need to write math.sin(foo) rather than just sin(foo), or
else change
 import math
to
 from math import sin

Has this code been tested? Please try to test any changes
to cwm before you check them in.

Hmm... it seems there's a test file in the regression
suite that's trying to run...

$ make
PYTHONPATH=`/bin/pwd`/.. python retest.py -c regression.n3
list/detailed.tests
online.tests norm/detailed.tests
  1/76 list/detailed.tests#t10394      Various trigo builtins
Traceback (most recent call last):
  File "../cwm.py", line 652, in ?
    doCommand()
  File "../cwm.py", line 495, in doCommand
    think(workingContext, mode=option_flags["think"])
  File "/home/connolly/w3ccvs/WWW/2000/10/swap/query.py", line 58, in
think
    return InferenceTask(knowledgeBase, ruleFormula, mode=mode,
repeat=1).run()  File "/home/connolly/w3ccvs/WWW/2000/10/swap/query.py",
line 206, in run
    return self.runSmart()
  File "/home/connolly/w3ccvs/WWW/2000/10/swap/query.py", line 195, in
runSmart    total += cy.run()
  File "/home/connolly/w3ccvs/WWW/2000/10/swap/query.py", line 301, in
run
    return rule.once()
  File "/home/connolly/w3ccvs/WWW/2000/10/swap/query.py", line 400, in
once
    total = query.resolve()
  File "/home/connolly/w3ccvs/WWW/2000/10/swap/query.py", line 541, in
resolve
    return self.unify(self.queue, self.variables, self.existentials)
  File "/home/connolly/w3ccvs/WWW/2000/10/swap/query.py", line 704, in
unify
    nbs = item.tryBuiltin(queue, bindings, heavy=0, evidence=evidence)
  File "/home/connolly/w3ccvs/WWW/2000/10/swap/query.py", line 980, in
tryBuiltin
    result = pred.evalObj(subj, queue, bindings.copy(), proof,
self.query)
  File "/home/connolly/w3ccvs/WWW/2000/10/swap/term.py", line 686, in
evalObj
    return self.store._fromPython(self.evaluateObject(subj.value()))
  File "/home/connolly/w3ccvs/WWW/2000/10/swap/cwm_trigo.py", line 70,
in evaluateObject
    return acos(numeric(subj_py))
NameError: global name 'acos' is not defined
diff: list/ref/trigo.n3: No such file or directory
Files differ, result= 512
######### from normal case detailed.tests_t10394.out: cwm
math/trigo-test.n3 --think --purge-rules
make: *** [all] Error 255


ok, fixed those problems... (v1.7)

Now I'm getting...

diff: list/ref/trigo.n3: No such file or directory

and I can't find it in CVS:

$ cvs update ref/trigo.n3
cvs server: nothing known about ref/trigo.n3

did you check it in?

eek! you seem to have clobbered lots of stuff
in detailed.tests
revision 1.5
date: 2003/09/25 22:28:30;  author: kdubost;  state: Exp;  lines: +4 -73

so I backed out to v1.4 of detailed.tests

But now I'm losing at this test...

13/87 norm/detailed.tests#norm10      Bug with RDF output in bySubject
mode

Looks like it's time to re-affirm the following agreement:

"Cwm developers agree that all the tests have to pass before code is
checked in. To run the tests, do make in the swap/test directory."
  -- http://www.w3.org/2000/10/swap/doc/cwm.html#dev


-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/

Received on Monday, 13 October 2003 18:01:42 UTC